China has warned about potential risks associated with Anthropic's Claude Code, an AI coding tool. The warnings reportedly concern a possible "backdoor" security risk that could compromise data. Anthropic has responded to the warnings, though the nature and validity of the concerns are unclear.